I am a little bit confused by why/when to smooth for game models.
If I want sharp edges on a cube that I am going to smooth I need to add edge loops close to the edges I want to stay sharp, right?
The question to me is - do I want that all the time or even most of the time? Why not just "not smooth" it?
Is the idea to do a low poly, smooth it (thereby making it high poly) and then generate a normal map from it? Or will I not need to retopo the smoothed version anyways? I did a Chesterfield sofa once from a tutorial where the unsmoothed version was still too high poly, so I had to manually retopo and then bake the normals. Or am I on the wrong track here altogether?
Do I want to smooth it so I can get into more detailed modelling? Smoothing is not exactly the same as subdividing, right? If I subdivide a cube in zbrush or mudbox it keeps its edges anyway, if I recall correctly.
Maybe someone can enlighten me on the involvment of smoothing and edge loops to maintain hard edges in game production ...
Thanks
Coffee
Replies
http://wiki.polycount.com/wiki/Normal_Map_Modeling#Smoothing_Groups_.26_Hard_Edges
Modeling for games is a fairly complex subject, but this might help:
http://wiki.polycount.com/wiki/Normal_Map_Modeling
Info about grabbing highpoly details:
http://wiki.polycount.com/wiki/Texture_Baking
Coffeehouse, you are on the right track. It's common to have a subD working model, a high poly model, and a low poly unwrapped model.
I think your confused by the term smooth. It does usually refer to subD (Max's TurboSmooth or Maya's Smooth
Mesh Preview). And just adding geometry without changing the form is called tessellating.